利用合成控制法估計加入歐盟對經濟成長的影響：來自中歐四國的證據 - 政大學術集成

全文

(1)國立政治大學應用經濟與社會發展英語碩士學位學程 International Master’s Program of Applied Economics and Social Development College of Social Sciences National Chengchi University 碩士論文. 立. Master’s 政治Thesis. 大. ‧ 國. 學 y. sit. Nat. 來自中歐四國的證據. ‧. 利用合成控制法估計加入歐盟對經濟成長的影響:. er. io. Estimating the Impact of Joining the European Union on. n. al v the Economic Growth n i Control Method: C Using Synthetic hengchi U. Evidence from the Visegrad Group Countries. Student: Lenka Pomykalová Advisor: Tzu-Ting Yang. 中華民國 109 年 06 月 June 2020. DOI:10.6814/NCCU202000585.

(2) 立. 政治大. ‧. ‧ 國. 學. n. er. io. sit. y. Nat. al. Ch. engchi. i Un. v. DOI:10.6814/NCCU202000585.

(3) Acknowledgement I would like to express my deepest gratitude to my thesis advisor, Professor Yang for his guidance and support and many helpful comments regarding my research. My further appreciation goes to my family for their support and making it possible for me to study abroad. Last but not least, my deepest thanks goes to all my friends for their love and support.. 立. 政治大. ‧. ‧ 國. 學. n. er. io. sit. y. Nat. al. Ch. engchi. i Un. v. DOI:10.6814/NCCU202000585.


(5) Abstract The main aim of the thesis is to estimate the effect of joining the European Union on the Visegrad group countries. The growth effects were estimated for the GDP per capita and the GDP components of the Visegrad group countries by using the synthetic control method. The results of the analysis show that the membership in the European Union had a positive effect on the GDP per capita of three countries, Czech Republic, Slovakia and for the last treated years, Poland. The results revealed negative effect of the membership on the GDP per capita in Hungary. The analysis also showed that the net export per capita was the main driving force of the GDP per capita growth in the Visegrad Group countries, gained by the increased trade with the European Union members and the single market accession. The results for the rest of the GDP. 政治大. components were inconclusive and mostly statistically insignificant.. 立. ‧ 國. 學. Keywords. European Union, Visegrad Group, Synthetic Control Method, GDP, Czech Republic,. ‧. Slovak Republic, Poland, Hungary. n. er. io. sit. y. Nat. al. Ch. engchi. i Un. v. DOI:10.6814/NCCU202000585.


(7) Table of Contents 1. Introduction .......................................................................................................... 1. 2. Background .......................................................................................................... 4 The European Union ........................................................................................ 4 The 2004 EU Enlargement .............................................................................. 6 Visegrad Group ................................................................................................ 8. 3. Literature Review .............................................................................................. 13. 4. Data Description................................................................................................. 17. 政治大 Variables ........................................................................................................ 19 立 Sample ........................................................................................................... 18. ‧ 國. 學. Summary statistics ......................................................................................... 20 Methodology ....................................................................................................... 24. 6. Preliminary results............................................................................................. 27. 7. Results ................................................................................................................. 35. ‧. 5. sit. y. Nat. io. al. er. Czech Republic .............................................................................................. 35 Hungary ......................................................................................................... 45. n. iv n C Poland ............................................................................................................ 56 hengchi U Slovak Republic ............................................................................................. 66 Results comparison ........................................................................................ 75 8. Discussion............................................................................................................ 81. 9. Conclusion .......................................................................................................... 85. 10. References ........................................................................................................... 88. Appendix A: Data description .................................................................................. 94 Appendix B: Detail of SCM results .......................................................................... 95. DOI:10.6814/NCCU202000585.

(8) List of Tables Table 1: List of donor pool countries ........................................................................... 18 Table 2: List of predictors ............................................................................................ 20 Table 3: Summary statistics of variables for Czech Republic ..................................... 21 Table 4: Summary statistics of variables for Hungary ................................................. 21 Table 5: Summary statistics of variables for Slovak Republic .................................... 22 Table 6: Summary statistics of variables for Poland .................................................... 22 Table 7: Summary statistics of variables for donor pool ............................................. 23 Table 8: Results of the SCM analysis for the treated countries ................................... 82 Table 9: Detail of covariates ........................................................................................ 94. 政治大 Table 11: Predictor balances, Czech Republic ............................................................. 95 立 Table 10: Detail of outcome variables ......................................................................... 94. Table 12: Estimated p-values for Czech Republic ....................................................... 97. ‧ 國. 學. Table 13: Predictor balances, Hungary ........................................................................ 99 Table 14: Estimated p-values for Hungary ................................................................ 101. ‧. Table 15: Predictor balances, Poland ......................................................................... 104. y. Nat. Table 16: Estimated p-values for Poland ................................................................... 106. io. sit. Table 17: Predictor balances, Slovak Republic .......................................................... 108. n. al. er. Table 18: Estimated p-values for Slovak Republic .................................................... 110. Ch. engchi. i Un. v. DOI:10.6814/NCCU202000585.

(9) List of Figures Figure 1: GDP per capita in the treated countries and EU15 ....................................... 27 Figure 2: Trade balances for the V4 countries ............................................................. 29 Figure 3: Government spending per capita in the treated countries ............................ 30 Figure 4: Private consumption per capita in the treated countries ............................... 31 Figure 5: Investment per capita in the treated countries .............................................. 32 Figure 6: Comparison of GDP components before and after the EU accession .......... 33 Figure 7: Comparison of GDP per capita, Czech Republic ......................................... 36 Figure 8: Comparison of export per capita, Czech Republic ....................................... 37 Figure 9: Comparison of import per capita, Czech Republic ...................................... 38. 政治大. Figure 10: Comparison of net export per capita, Czech Republic ............................... 39 Figure 11: Comparison of private cons. per capita, Czech Republic ........................... 40. 立. Figure 12: Comparison of gov. spending per capita, Czech Republic ......................... 41. ‧ 國. 學. Figure 13: Comparison of investment per capita, Czech Republic ............................. 42 Figure 14: Robustness check, Czech Republic ............................................................ 43. ‧. Figure 15: Comparison of GDP per capita, Hungary................................................... 46 Figure 16: Comparison of export per capita, Hungary ................................................ 47. y. Nat. sit. Figure 17: Comparison of import per capita, Hungary ................................................ 48. al. er. io. Figure 18: Comparison of net export per capita, Hungary .......................................... 49. n. Figure 19: Comparison of private cons. per capita, Hungary ...................................... 50. Ch. i Un. v. Figure 20: Comparison of gov. spending per capita, Hungary .................................... 51. engchi. Figure 21: Comparison of investment per capita, Hungary ......................................... 52 Figure 22: Robustness check, Hungary ....................................................................... 53 Figure 23: Comparison of GDP per capita, Poland ..................................................... 57 Figure 24: Comparison of export per capita, Poland ................................................... 58 Figure 25: Comparison of import per capita, Poland ................................................... 59 Figure 26: Comparison of net export per capita, Poland ............................................. 60 Figure 27: Comparison of private cons. per capita, Poland ......................................... 61 Figure 28: Comparison of gov. spending per capita, Poland ....................................... 62 Figure 29: Comparison of investment per capita, Poland ............................................ 63 Figure 30: Robustness check, Poland .......................................................................... 63 Figure 31: Comparison of GDP per capita, Slovak Republic ...................................... 67. DOI:10.6814/NCCU202000585.

(10) Figure 32: Comparison of export per capita, Slovak Republic .................................... 68 Figure 33: Comparison of import per capita, Slovak Republic.................................... 69 Figure 34: Comparison of net export per capita, Slovak Republic .............................. 70 Figure 35: Comparison of private cons. per capita, Slovak Republic .......................... 71 Figure 36: Comparison of gov. spending per capita, Slovak Republic ........................ 72 Figure 37: Comparison of investment per capita, Slovak Republic ............................. 73 Figure 38: Robustness check, Slovak Republic ........................................................... 73 Figure 39: Difference between the outcome variables and the synthetic estimates ..... 80. 立. 政治大. ‧. ‧ 國. 學. n. er. io. sit. y. Nat. al. Ch. engchi. i Un. v. DOI:10.6814/NCCU202000585.

(11) List of Abbreviations CZ. Czech Republic. EU. European Union. EU10. Countries that joined the European Union in 2004. EU15. Fifteen European Union member states before 2004. Eurozone. Monetary union of countries that adopted euro. FDI. Foreign direct investment. GDP. Gross domestic product. HUN. Hungary. PL. Poland. SCM. Synthetic Control Method. SK. Slovak Republic (Slovakia). V4. Visegrad Group. 學 ‧. ‧ 國. 立. 政治大. n. er. io. sit. y. Nat. al. Ch. engchi. i Un. v. DOI:10.6814/NCCU202000585.


(13) Introduction. 1. Introduction. Sixteen years ago, the European Union (EU) experienced its largest enlargement ever since its establishment, when ten countries joined the organization. The enlargement connected the Western and Eastern Europe. The GDP per capita level of the newly accepted EU countries was much lower in comparison to the old member states. Moreover, the mainly politically oriented enlargement was long planned, as the candidate countries had to implement the EU legislation and regulations. Ever since, the EU had been challenged many times by economic, political and migration crises. Those led many to question the importance of the organization and whether the member countries benefit from the economic integration. Notably, ever since the EU was established, there was no war conflict on the continent, which can be granted to the. 政治大 their own country’s sovereignty 立 and independence.. institution. Nonetheless, many EU citizens seem to be in favor of leaving the EU to gain. ‧ 國. 學. To question the EU importance and its effect on the member states, the thesis aims to estimate the EU participation impacts on the GDP per capita and its components for the Visegrad Group (V4) countries. The V4 was founded after the collapse of the. ‧. Soviet Union by the following countries, Czech Republic, Slovak Republic (Slovakia),. sit. y. Nat. Hungary and Poland, to cooperate during their market transition. The countries share. io. er. similar history and geographical position. With their tight relations, the V4 countries create an intriguing sample to test the EU participation effect. While Poland, in terms. n. al. i Un. v. of territory, is the sixth largest in the EU and the first one in V4, with relatively low. Ch. engchi. value of export in the GDP and large internal market, the other three V4 countries are small, trade-oriented economies (European Union, n.d.). V4 countries joined the EU on 1 May 2004, which provides sufficient post-treatment period. Their GDP per capita levels were much lower compared to the old member states. Based on the income convergence theory, countries with lower income per capita levels reach higher economic growth rate than the richer economies and in the long run, they will reach the income per capita level of the richer countries. Notably, the current research focusing on the EU and its effect on the member countries shows mainly positive results, but the authors often address fragility of their estimates due to the methods used to estimate the effects. The lack of credibility is assumed to the omitted variables, causality concerns, heterogeneity, or measurement errors. In order to broaden the past research, the thesis uses the synthetic control method. 1. DOI:10.6814/NCCU202000585.

(14) Introduction. (SCM) to estimate the economic impact on the V4 economies. SCM is strictly data-driven method that generates transparent results and compared to other methods, avoids extrapolation, and usually provides results with higher credibility. The main aim of the thesis is to estimate the economic benefits from EU participation on the V4 countries. The synthetic estimated units that have never joined the EU were built from a donor pool of non-EU countries by using the SCM. Additionally, the thesis aims to estimate the EU participation effect on the GDP per capita components to clarify the main driving force behind the EU participation effect. The main goal of the thesis is to answer the three following research questions. RQ1: What is the effect of joining the EU on the GDP per capita levels of the V4 countries? RQ2: What is the effect of the EU participation on the GDP components of the V4 countries? RQ3: What GDP component is the main driving force of the GDP per capita growth?. 立. 政治大. The results estimated in this thesis showed that in the Czech Republic and Slovak. ‧ 國. 學. Republic, the GDP per capita levels increased after the EU accession. Poland also received positive effect of the EU accession in the last treated years. For Hungary, the. ‧. obtained effect from the EU participation on the GDP per capita was negative, as due to the Hungary’s economic situation. The analysis performed illustrated that joining the. Nat. sit. y. EU as itself is not a key to a strong economic growth, and the potential members of the. io. er. EU firstly need to have a strong and healthy economy. Latter analysis revealed net export per capita is the main economic benefit for the new EU countries. The results. n. al. Ch. i Un. v. showed strong growth of the net export per capita for all the V4 countries. For the rest. engchi. of the GDP components, the results were not unanimous and were mostly statistically insignificant. The income per capita for the V4 group members shows relative convergence to the old EU member countries, and therefore, the research confirms the income convergence theory. Furthermore, the thesis contributes to the past research by detailed analysis of the V4 countries and their economic benefits/losses from the EU participation. The estimates also present the main benefits from joining the EU and illustrate the EU participation effects on each of the GDP components, which was not previously investigated. The thesis confirms the EU participation positively influences the GDP per capita growth of the member states, however, the countries need to be economically ready to join the organization to fully generate the income per capita increase.. 2. DOI:10.6814/NCCU202000585.

(15) Introduction. The structure of the thesis is as follows. The background review discusses the history of the EU, the V4 economic transition and each of the countries’ economic environment. The literature review covers the previous research and the methods used for estimating the economic benefits from the EU membership. The latter two chapters describe the data sample, variables and the methodology used in this thesis. Chapter six covers preliminary analysis of the EU effects and the main results of the SCM analysis are in chapter seven. The next chapter discusses the results and suggestions for the potential members and the member countries. The last chapter contains conclusion, limitations of the study and future research suggestions.. 立. 政治大. ‧. ‧ 國. 學. n. er. io. sit. y. Nat. al. Ch. engchi. 3. i Un. v. DOI:10.6814/NCCU202000585.

(16) Background. 2. Background. This chapter discusses the EU history, the achievements in the economic integration of the EU member countries and the accession process. The 2004 enlargement is emphasized. Furthermore, this chapter describes the V4 history, its purpose, the economic situation of the member countries and their achievements from joining the EU.. The European Union After the Second World War, the European countries were economically devastated and wanted to prevent any new war conflict on the continent. A French economist Jean. 政治大. Monnet introduced an idea to integrate the European countries, which was later presented by the French foreign minister Robert Schuman (Baldwin, Wyplosz &. 立. Wyplosz, 2006). Schuman proposed to create an organization that would prevent any. ‧ 國. 學. further war conflicts between Germany and France by pooling their supplies of coal and steel together and afterwards, the European Coal and Steel Community was created. ‧. (Baldwin et al., 2006). The EU history based on Monnet’s idea started in 1951 by signing the Treaty of Paris1, which guaranteed peace by gathering the coal and steel. Nat. sit. y. supplies of Germany and France and another four European countries, Belgium, the. io. er. Netherlands, Luxembourg and Italy, that decided to join the community (Baldwin et al., 2006). Coal and steel were essential products for the war industry and rebuilding. n. al. Ch. i Un. v. military forces, hence cooperation in this field was believed to restore peace on the. engchi. continent. The treaty was supposed to last 50 years; however, it was replaced before its expiration by higher degree of economic integration (Baldwin et al., 2006). In 1957, after experiencing the well-maintained cooperation, the founding countries decided to step further in the economic integration. Two treaties were signed in Rome by the member countries. The treaties firstly established the European Economic Community (EEC) and secondly, the European Atomic Energy Community (Euratom)2, both treaties were signed for an unlimited amount of time (Dedman, 2006). The EEC created customs union that implied a united tariff on imports conducted outside of the EEC and removed the tariffs and quotas inside the EEC. Furthermore, the. 1 2. Formally the Treaty establishing the European Coal and Steel Community. The Euratom treaty aimed to integrate the nuclear industry between the member countries.. 4. DOI:10.6814/NCCU202000585.

(17) Background. EEC treaty established a common market with labor, people, capital and service mobility (Baldwin et al., 2006). The former treaties were in 1992 replaced by the Treaty on the European Union3, which created the EU and change the name from the European Community to the current name, European Union (Baldwin et al., 2006). The treaty brought unified foreign policy, security policy and common internal security measure (Archick, 2005). The treaty created economic and monetary union and thus, single currency was created in 1999 and replaced most of the national currencies in 2002 (Baldwin et al., 2006). The EU currency is currently used by all the member countries, except for Denmark, Sweden, Bulgaria, Croatia, Czech Republic, Hungary, Poland and lastly, Romania. The community firstly expanded in 1973, when three countries decided to join. Those countries were the United Kingdom, Denmark and Ireland (Baldwin et al., 2006).. 政治大. Further on, in 1980s, Greece, Portugal and Spain decided to join (Baldwin et al., 2006).. 立. Moreover, Austria, Finland and Sweden finally joined the EU in 1995 and the number. ‧ 國. 學. of EU countries increased to fifteen (Baldwin et al., 2006). The following 2004 enlargement was the biggest ever since the establishment of the EU and extended the. ‧. number of member countries to 25. In the following years, Bulgaria, Romania and Croatia joined and the EU expanded to 28 members. Currently, there are officially only. Nat. sit. y. 27 EU members as the United Kingdom decided to leave the EU on 31 January 2020.. io. er. Nowadays, the EU is based on deep economic integration. The member countries enjoy all four freedoms of the single market4, some also use the common currency and. n. al. Ch. i Un. v. joined the Schengen Area that allows free pass across the borders. The main economic. engchi. benefits of the EU are clear. The member countries benefit from trade with no barriers. Firms in the EU can easily expand their businesses in the EU area with significantly reduced costs and furthermore, the member countries have access to the funding from the EU budget. Moreover, the big size of the EU allows for many opportunities in some important areas, such as investment or R&D (Kutan & Yigit, 2007). Nonetheless, the EU integration cannot be achieved without any cons, such as high level of bureaucracy and regulations. Indeed, collaboration between 27 countries with different languages and cultural background requires high bureaucratic burden. Furthermore, participation in the EU brings costs to every country as each. 3. Also called the Maastricht Treaty. The treaty went into effect on November 1, 1993 (Archick, 2005). The four freedoms of the single market are the free movement of labor, capital, services and lastly, goods. 4. 5. DOI:10.6814/NCCU202000585.

(18) Background. of the members must transfer money to the EU budget. The single market allows for free mobility of people between the states. As there are big differences in the GDP levels in the EU, those countries with higher economic development might see immigration increase from less developed states as a threat. Lastly, by joining the EU, the member countries also agree to join the eurozone 5 , which limits their relative monetary independence. Certainly, the EU integration process is complex and problematic. While some countries may want to achieve higher degree of integration, some may opt for a lower one.. The 2004 EU Enlargement The 2004 EU enlargement6 was the biggest ever since the EU was established. The. 政治大. newly admitted countries had to undertake a transition towards liberal markets before the EU accession. Besides, those countries were mainly not as developed as the previous. 立. EU countries, and their production was usually more labor oriented with lower level of. ‧ 國. 學. productivity.. The 2004 enlargement consisted of ten countries, Czech Republic, Cyprus,. ‧. Estonia, Hungary, Latvia, Lithuania, Malta, Poland, Slovakia and Slovenia (EUR-Lex, 2007-a). The formal negotiations regarding the EU accession started in 1998 with the. y. Nat. sit. Czech Republic, Cyprus, Estonia, Hungary, Poland and Slovenia, in the following year,. al. er. io. the negotiations were initiated with Bulgaria, Latvia, Lithuania, Malta, Romania and. v. n. Slovakia (Archick, 2005). Before the 2004 enlargement, the EU consisted in total of 7. fifteen countries .. Ch. engchi. i Un. Before the official accession, the potential EU members must meet the Copenhagen criteria that consist of three main parts, firstly the countries must have stable democracy, protection of human rights and rights of minorities, secondly functioning market economy with sufficient competition level and lastly, obligation to join the union and follow its political, economic and monetary aims (EUR-Lex, n.d.). Once all rules and regulations are met and the negotiations are done, the process of accession treaty ratification starts, the treaty needs to be signed firstly by the EU institutions and also by all EU member states, which in total can take up to two years 5. Eurozone is a monetary union of the EU countries that adopted euro and use it as their national currency. Often referred to as the Eastern enlargement. 7 Those fifteen countries were United Kingdom, France, Germany, Italy, Belgium, Greece, Netherlands, Luxembourg, Denmark, Spain, Portugal, Ireland, Sweden, Finland and Austria. They are often referred to as the EU15. 6. 6. DOI:10.6814/NCCU202000585.

(19) Background. (Archick, 2005). In 2002, on a summit in Copenhagen, EU countries terminated the accession process with the ten candidate countries. The ten candidate countries officially joined the EU on 1 May 2004 (Archick, 2005). The 2004 EU enlargement was the biggest in the EU history and it brought diversity into the EU. The EU population increased by nearly 20% due to the accession by the ten new member countries (Crespo Cuaresma et al., 2008). Thus, the EU’s population increased to approximately 450 million citizens (Archick, 2005). The GDP of the newly admitted countries was mainly lower compared to the old member countries and thus, the accession was rather politically than economically oriented. The 2004 enlargement closed the gap in democracy between the East and West 8 (Sedelmeier, 2014). According to Zielonka (2004), the candidate countries’ GDP altogether was less than 5% of the EU15 GDP level. The large gap in the welfare. 政治大. between the EU10 and the previous member states caused a certain level of disruption. 立. in the single market. In Eastern markets, the labor productivity was much lower and. ‧ 國. 學. agriculture share was much higher compared to the EU15 (Baldwin et al., 1997). The economic benefits for the ten new EU countries were obvious. As per report. ‧. done two years after the enlargement, the new member countries (EU10) experienced a significant economic growth of 3.75% on average (EUR-Lex, 2007-b). The main. Nat. sit. y. contribution in higher income growth is assumed to the increased trade with the EU. io. er. countries. Moreover, the new EU countries also experienced a large increase in foreign direct investment (FDI) and foreign firms allocating their branches in the territories of. n. al. Ch. i Un. v. the new member countries (EUR-Lex, 2007-b). Joining the EU also improved the global. engchi. image of the new EU countries. Joining a large economic organization improved their role in the global market and instigated investors to allocate their investments to their economies. Lastly, joining the EU also improved their labor and pension systems. The old member states had to face some threats connected with the large enlargement. One of the main threats was large migration due to the significant economic differences between the old EU15 and new member states. As the new countries joined the single market, the labor migration became more possible than ever. Nonetheless, the new EU countries had to wait to join the Schengen area. Even though the V4 countries signed the Schengen agreement in 2003, they were allowed to join the Schengen zone on 21 December 2007 (Schengen Visa Information, n.d.).. 8. Nonetheless since year 2010, the democratic quality in Hungary dropped (Sedelmeier, 2014).. 7. DOI:10.6814/NCCU202000585.

(20) Background. Lastly, the budget of the EU was affected by the new member states. As those were mainly countries with low GDP levels, the EU budget was mostly used to provide aid to the newly admitted states. Indeed, some of the countries from the 2004 enlargement, such as the V4 countries, still receive more aid from the EU budget than they contribute, which makes them so called net recipients.. Visegrad Group The V4 was created by four countries that wanted to cooperate during the market reforms process and on joining international organizations. The satellite states had to liberalize their markets, open the economies to the world, privatize the former nationally owned firms and tackle down the high debts resulted from the previous governments.. 政治大. Currently, the V4 consists of four member countries, Czech Republic, Hungary, Slovak Republic and Poland. V4 group was founded in 1991 by signing the Visegrad. 立. Declaration and it aimed to fulfill two main objectives, firstly the economic transition. ‧ 國. 學. after the era of the Soviet Union and secondly, on joining two important organizations, EU and NATO (Dangerfield, 2009). Both were achieved, nonetheless, the V4 still. ‧. exists, and currently aims to promote cooperation between the V4 regions to work. sit. Nat. mainly foreign policy and regional activities (Dangerfield, 2009).. y. together. The V4 maintains cooperation between its member states in several key areas,. al. er. io. All four V4 countries joined the EU in 2004 as relatively poor in comparison to. v. n. the old EU member countries. Currently, the V4 group countries are net receivers as the. Ch. i Un. money they get back from the EU budget exceed their contributions. The accession to. engchi. the EU improved the economic growth of the V4 countries via accession to the EU market and increased trade inside the EU single market. Following on the income convergence theory, the V4 countries as relatively poorer are expected to converge with their income levels to those old EU member countries. Therefore, their income per capita growth is expected to increase after joining the EU. The EU accession was expected to bring the V4 trade liberalization, FDI, skills, technology, and capital that will boost up the economic development (Tang, 2000). Furthermore, the countries liberalized their previously monopolized markets, such as energy, telecommunication or aviation and benefited from the common agricultural policy (Jedlička, Kotian & Munz, 2014).. 8. DOI:10.6814/NCCU202000585.

(21) Background. 2.3.1. Economic History of V4 Countries. In order to interpret the results of the EU accession, the economic history of the studied countries is briefly covered. As was previously mentioned, the V4 countries share similar background of being satellite states of the Soviet Union and this affected their economies greatly by high level of debt, low trade openness and low competition levels. Most of the centrally planned economies crashed in 1989 with the collapse of the Berlin Wall and afterwards, their governments faced resulting issues, such as need for privatization and radical changes that mostly led to fall in recession (Berend & Berend, 2009).. The. studied. countries. started. their. economic. transition. in. 1989. (Koyame-Marsh, 2011). Baldwin (1994) indicated the V4 countries as those who may enter the EU first.. 治政大rich countries, Poland was much Czech Republic and Hungary were small and relatively 立 bigger and poorer, and so was Slovakia that had even lower GDP per capita level. He also highlighted the differences in the economic maturity of the countries. While. ‧ 國. 學. (Baldwin, 1994). Even though the V4 countries agreed to adopt the single currency, Slovakia was the only one to join the eurozone in 2009. The V4 countries achieved. ‧. significant development after their market liberalization. Currently, Poland receives the highest amount of money from the EU budget (Kovacevic, 2019). According to Ivanová. y. Nat. sit. and Masárová (2018), in terms of competitiveness of the countries, Slovakia reaches. al. er. io. the lowest levels, while the Czech Republic achieved the highest level, followed by. v. n. Poland and Hungary, respectively. During the world economic crisis, only Poland. Ch. i Un. managed to avoid recession (Ivanová & Masárová, 2018).. engchi. Czech Republic The above-mentioned country was officially found in 1993, when the Czechoslovakia split into two countries. After 1989 and the Velvet Revolution, the country had gone through major economic changes, including privatization and economic reforms. The liberalization of Czech economy started with price liberalization and in 1991, Czech Republic started coupon privatization and opened the domestic market to international trade. The transition in the Czech Republic was successful and turned Czech Republic into a strong export-oriented economy (Myant, 2007). During the two decades after transition, real GDP in Czech Republic grew approximately by 37% (Koyame-Marsh, 2011).. 9. DOI:10.6814/NCCU202000585.

(22) Background. Overall, the Czech Republic is extremely export oriented country with its trade openness 9 exceeding 100%. The impact of the global economic crisis on the Czech Republic was not as large as in comparison to the other EU countries, especially due to the ability to perform monetary policies. Currently, Czech economy is the richest in terms of GDP per capita out of the studied countries. From 1991 to 2017, the level of GDP per capita was well above the rest of the V4 countries, nonetheless, it was still way below the average GDP per capita of the old fifteen EU states.. Hungary After 1989, Hungary experienced a large fall in export to the Eastern countries and faced large foreign debt, deficit of the public finance and high unemployment rate, which soon caused the economy to fall to a recession (Žídek, 2014). Four years later,. 政治大. Hungary received a loan from the IMF and assistance to reduce the deficits in public. 立. finance and trade balance (Réti, 1995). Soon after the reforms, Hungary experienced. ‧ 國. 學. inflow of FDI, increase of export, and the GDP per capita level recovered (Žídek, 2014). Just as the previous country, Hungary is export oriented country with the trade. ‧. ratio in some years exceeding 100% of the Hungary’s GDP (Pintér, 2018). Compared to other post-Soviet countries, Hungary received high value of FDI per capita (Ungváry,. sit. y. Nat. 2018). Furthermore, Hungary was determined as the most promising EU candidate. io. er. (Böröcz, 2012). Nonetheless, when Hungary joined the EU, it was not able to perform economic structural reforms as other V4 countries did and its economy faced large. n. al. i Un. v. domestic and foreign deficit (Beacháin, Sheridan & Stan, 2012). This probably. Ch. engchi. influenced its benefits from joining the EU.. During the global economic crisis in 2008, Hungary experienced recession due to the high level of government debt and increase in unemployment that exceeded 10% (Žídek, 2014). After this, Hungary had to receive a rescue package from the IMF. Furthermore, after the global crisis, Hungary’s GDP per capita fell sharply and the economy did not fully recover.. 9. Trade openness is defined as a ratio of trade (exports plus imports) to GDP.. 10. DOI:10.6814/NCCU202000585.

(23) Background. Poland Among the V4 countries, Poland is the largest, however, least developed. After 1989, Poland faced large foreign debt and hyperinflation (Gomułka, 2016). The situation in Poland was much worse compared to the other V4 countries and the government decided to implement a shock therapy reform called the Balcerowicz Plan introduced in the beginning of 1990 that implied immediate market liberalization and currency devaluation, however this also led to recession and hyperinflation (Žídek, 2011). In the following years, Poland managed to tackle down the inflation and reach surplus in government budget (Ratajczak, 2009). After the privatization until 2013, the Polish GDP was experiencing approximately growth of 4% annually (Gomułka, 2016). Just as other post-soviet countries, Poland’s export was lacking competitiveness.. 治政大 however, the economic growth during EU membership was never negative 立 (Kolodziejczyk, 2016). Poland experienced large increase in exports, FDI inflow and The positive economic results from joining the EU took a relatively long period,. ‧ 國. 學. improved its productivity (Kolodziejczyk, 2016). Nonetheless, the living standards in Poland are still very low compared to the western EU countries. According to. ‧. Kolodziejczyk (2016) the nominal output level in Poland reached the 8th highest value from the EU countries, however, with a large gap. Poland experienced large amount of. y. Nat. sit. migration. The Central Statistical Office in Poland estimates the amount of Polish. al. er. io. citizens working abroad to increase by almost one million between 2004 and 2013. v. n. (Kolodziejczyk, 2016). Furthermore, during EU membership, Poland experienced. Ch. i Un. decrease in inequality and if it did not join the EU, its GDP would be probably much. engchi. lower (Kolodziejczyk, 2016). Nonetheless, compared to the EU15, Polish economic development is incredibly low. The large migration of Polish workers can be considered a negative consequence of the EU membership for Poland (Kolodziejczyk, 2016). Poland is one of few countries that did not suffer from the global economic crisis in 2008. The credit is assumed to large market and friendly and liberal environment. During the years before the crisis, Poland had stable inflation, output level and very low current account deficits (Belka, 2013).. Slovak Republic During the privatization, Slovakia had to overcome the consequences of the centrally planned economy, such as public finance deficit and high unemployment. It introduced. 11. DOI:10.6814/NCCU202000585.

(24) Background. critical economic reforms. Both, Slovakia and Czech Republic experienced in 1991 recession (Koyame-Marsh, 2011). With the price liberalization, the inflation rose sharply and was fought by monetary policy (Koyame-Marsh, 2011). After the disintegration of Czechoslovakia, Slovakia decided to switch from coupon privatization to auction, which caused that Slovak companies received large inflow of investment from foreign buyers (Koyame-Marsh, 2011). Moreover, Slovakia was mainly left with heavy and arms industries, which could not had been successfully traded with the EU countries (Koyame-Marsh, 2011). Slovakia also experienced high level of unemployment in early 1990s. During the transformation, Slovakia reached lower inflation and higher economic growth compared to Czech Republic. Due to the 2004 enlargement, Slovakia received a great number of foreign investments, lowered its inflation level, and reached higher economic growth.. 政治大. Moreover, with the EU membership, Slovakia experienced a large increase in export. 立. of goods and services. Slovak Republic is the only V4 country to adopt the EU common. ‧ 國. 學. currency in 2009. After joining the EU, Slovakia achieved high economic growths and its GDP per capita largely increased, converging to the level of the Czech Republic.. ‧. When the economic crisis hit the European countries, Slovakia experienced a large increase in unemployment and the GDP growth slowed down. Between 1989 and 2009,. Nat. n. al. er. io. sit. y. real GDP grew by approximately 60% (Koyame-Marsh, 2011).. Ch. engchi. 12. i Un. v. DOI:10.6814/NCCU202000585.

(25) Literature Review. 3. Literature Review. This chapter summarizes the past research conducted on analyzing the effect of the EU accession on the economic growth of the member states. Furthermore, the differences in the method and variables used in this thesis compared to the previous research are discussed. The previous literature on the effect of joining the EU on member countries is still insufficient. While there is a large number of research papers estimating the effect of the euro adoption on the member countries, there is only a limited amount of econometric research papers estimating the monetary benefits of the EU membership on the member countries. Furthermore, most of the previous research focuses on the past enlargements10. Also, no previous research solely focused on the V4 countries and their. 政治大 Nonetheless, based 立 on past research, the EU participation mostly led to higher. GDP and its components.. ‧ 國. 學. GDP per capita for the member countries (Badinger, 2005; Kutan & Yigit 2007; Crespo Cuaresma, Ritzberger-Grünwald & Silgoner, 2008; Campos, Coricelli & Moretti, 2014). The past literature mostly uses the Solow model or the endogenous growth. ‧. theory. The Solow model builds on an exogenous growth theory, where the long-term. sit. y. Nat. economic growth is achieved by the exogenous rate of technological change (Solow,. io. er. 1956). This means that either economic policy or integration would lead only to a level effect caused by temporarily higher economic growth rate. The Solow model introduced. n. al. i Un. v. the diminishing returns to capital, according to which, those countries that reached. Ch. engchi. lower output per capita will experience higher growth of the output level compared to those with higher values and in the long run, they would converge to the richer old member states. On the contrary, the endogenous growth theory accounts the technological change as an endogenous variable by firms investing to research to reach higher technological level (Romer, 1990). For this theory, the economic integration may cause a long-run positive effect on the economic growth of the country. Furthermore, the profits generated by higher technological levels by investing to research and innovation encourages the long-run economic growth, and the long-run economic growth boosts up with the larger size of the economy (Crespo Cuaresma et al., 2008).. 10. EU enlargements that took place before year 2004.. 13. DOI:10.6814/NCCU202000585.

(26) Literature Review. There is not a clear consensus what theory should be implemented, nonetheless, most researchers choose to use the Solow model due to its relative simplicity. Overall, the results of the previous research are mostly consistent, agreeing on the positive effects of participation in the EU on most of the countries, except for Greece that experienced negative effect. One of the first researches on the membership effect was by Henrekson, Torstensson and Torstensson (1997), the authors found significant positive effect on the economic growth of the member countries. Nonetheless, the authors warned against not completely robust data to change of control variables, and also measurement errors. Furthermore, Badinger (2005) found positive impacts of the EU membership on the income per capita of fifteen member states, however, the results were not completely robust. Badinger (2005) estimated by panel data regression that the sum of the EU. 政治大. members’ income per capita would be lower by approximately one-fifth without the. 立. economic integration. The estimated scale growth effect was only temporary; however,. ‧ 國. 學. the level effects were sizable (Badinger, 2005). Similar robust results were retrieved by Böwer and Turrini (2010), the authors estimated by using panel regression increased. ‧. income per capita growth generated after the EU accession.. Kutan and Yigit (2007) also estimated positive results generated by the EU. Nat. sit. y. participation on the EU members that occur especially in the long run. Furthermore, the. io. convergence levels of the old EU15 (Kutan & Yigit, 2007).. n. al. Ch. er. authors estimated positive effect of the EU accession on the output, productivity and. i Un. v. Furthermore, Crespo Cuaresma et al. (2008) used panel regression to estimate the. engchi. membership effects, and the authors determined that the EU accession has a positive, nonetheless, unbalanced effect on the long-term economic growth of the EU members. The effect was higher for the relatively poorer countries 11 , which confirms the convergence theory (Crespo Cuaresma et al., 2008). While previous papers considered only the old fifteen member states, the following authors were also focusing on the Eastern enlargement. Based on the Solow growth theory, Mann (2015) found EU had small, however, positive medium-run impact on the economic growth of the new member countries led by the trade integration in the single market with substantial benefits that were not measured, such as higher attractiveness for investors and lower risk premium. Molendowski (2015) found that the effects of 11. The authors estimated the results only for the fifteen old member states, therefore considered only enlargements before year 2004.. 14. DOI:10.6814/NCCU202000585.

(27) Literature Review. joining the EU on the V4 economies slightly differed. While Poland received the highest GDP growth rate during the first ten years of the EU participation, Hungary came out with the worst results (Molendowski, 2015). European Commission (2001) scenario for the Eastern enlargement was that the EU10 economies can achieve as high as 5.5% growth rate due to the EU accession generated by FDI inflow and higher labor force growth due to the higher labor force participation. Rapacki and Próchniak (2008) based their study on the Solow model theory and tested the convergence levels of the EU10 and EU15. Rapacki and Próchniak (2008) estimated a significant positive effect from the EU accession on the economic performance of the Eastern enlargement countries encouraged by the FDI inflow, structural reforms and structural funds money inflow. Furthermore, the authors. 政治大. predicted that the convergence between the old EU15 and EU10 can take between eight. 立. and thirty-three years (Rapacki & Próchniak, 2008).. ‧ 國. 學. More results were presented by Breuss (2001) that estimated the EU10 will gain from the EU membership approximately ten times higher increase of the real GDP per. ‧. capita compared to the old EU15. Breuss (2001) estimated that Poland and Hungary will increase their real GDP the most, approximately by 8 to 9% a year, nonetheless the. Nat. sit. y. Czech Republic will gain less, approximately 5 to 6% a year.. io. er. Furthermore, Maliszewska (2009) estimated the benefits for the Eastern enlargement countries from joining the single market, the author expected increase of. n. al. Ch. i Un. v. wages in the new EU countries and in terms of GDP, Poland is expected to generate. engchi. 3.4% growth of GDP and for Hungary the gain is estimated to be 7%. The benefits are assumed to the trade liberalization, work mobility and elimination of the trade barriers (Maliszewska, 2009). By using the SCM, Campos et al. (2014) found that joining the EU had a positive effect on the member countries and furthermore, the authors estimated positive effects of joining the EU on most of the old fifteen member countries, except for Greece. The researchers estimated that if the member states never joined the EU, on average, their level of GDP per capita would be lower by 12% (Campos et al., 2014). Nonetheless, when estimating the effects for the Eastern enlargement, Campos et al. (2014) did not reach unanimous results. For year 2004 as a treatment, the authors found positive effects of the EU entry for the Czech Republic and opposite results were retrieved for the rest of the V4 countries. When the authors used 1998 as a treatment year, the effects turned 15. DOI:10.6814/NCCU202000585.

(28) Literature Review. positive for all treated countries, but Slovakia. However, by changing the treatment date, the authors were left with relatively short pre-intervention period, which may have affected the SCM results12. Overall, the authors found positive effect gained by joining the EU on the output levels of the member countries. Nonetheless, the authors did not discuss the results regarding the economic environment of the treated countries, which is an important factor that needs to be considered for the results interpretation. Most of the previous research warns against the credibility of the results13. Using SCM may prevent some of the difficulties with estimating the EU membership effect, therefore should present more accurate results than other methods, such as regression or Difference in Differences. By using the SCM, values that sum up to 1 are assigned to each of the countries, based on the similarities of the chosen predictors. The thesis complements the current research of the EU membership by estimating. 政治大. the results on the GDP per capita and the GDP components of four EU countries. The. 立. selected covariates are based on the previous research, but, complemented by several. ‧ 國. 學. different variables. Furthermore, the choice of the predictors is constant for each of the countries and outcome variables to ensure clear results driven by the same covariates.. ‧. n. er. io. sit. y. Nat. al. Ch. engchi. i Un. v. 12. See Abadie and Gardeazabal (2003). The researchers mainly warn against the results credibility due to lack of robustness of the results, heterogeneity, possible measurement errors and spill-over effects. 13. 16. DOI:10.6814/NCCU202000585.

(29) Data Description. 4. Data Description. This chapter presents the dataset and the sample used for the estimation. The thesis aims to analyze the impact of joining the EU on the V4 countries and therefore, the macroeconomic dataset is composed of standard economic growth predictors. Two data sources were used, World Bank’s Databank and United Nations Development Programme’s Human Development Report. Further explanation of the variables and the sources of the variables is attached in Appendix A: Data description. The selection of the outcome variables and predictors was based on Abadie, Diamond and Hainmueller (2015) complemented by several new predictors. The country-level data are collected for the period 1991-2017, depending on the country’s data availability14. The approximately decade long pre-treatment period should retrieve. 政治大 The year of accession, 2004, 立 or the year of signing the accession documents, 1998. satisfactory results. In the previous research were used two possible years as a treatment.. ‧ 國. 學. or 1999. The latter option provides too short pre-treatment period, which may lower the credibility of the results, therefore year 2004 was used as a treatment. The countries share the same donor pool and predictors to maintain the same conditions and. ‧. environment for each of the countries15.. sit. y. Nat. The credibility of the results is checked by three robustness tests16. First robustness. io. er. check is conducted by adding Iceland, Norway and Switzerland to the dataset. Those countries were firstly omitted from the donor pool as having signed trade agreements. n. al. i Un. v. with the EU. The second check was conducted by leaving-one-out re-analysis of the. Ch. engchi. sample, based on Abadie (2019). For each of the SCM estimates, the country with highest value assigned was dropped from the dataset. The third robustness check tests the robustness of the estimates on larger donor pool that consists of 37 countries17. The results of the robustness checks can be found at the end of each country’s results section18. 14. For the Czech Republic and Hungary, the data are retrieved from 1991, for Slovakia from 1992 and for Poland from 1994. This gives 13 years, 12 years and 9 years long pre-treatment panel, respectively. 15 The pre-intervention period length differs for the treated countries due to different data availability for each of the countries. 16 A robustness check proposed by Abadie (2019) with backdating the treatment was not used as the pre-treatment period for this check is not long enough. 17 The added countries for this robustness check are United States, Guatemala, Indonesia, Mexico, Mongolia, South Africa, Iceland, Norway and Switzerland. 18 In the graphs, robustness check of leaving one country out of the dataset is called "leave-one-out", for adding the three countries with EU trade agreements, the variable is called "add" and for larger donor pool, the robustness check is called "different donor pool".. 17. DOI:10.6814/NCCU202000585.

(30) Data Description. Sample Donor Pool Out of the world-level data were selected for further analysis only those with enough data and no missing observations for the outcome variables. Those that joined the EU in the pre-treatment or post-treatment period had to be omitted. Following previous methodology introduced by Böwer and Turrini (2010), those countries with large dependency on oil production and least-developed countries were omitted from the dataset. Furthermore, those countries for which the total sum of the GDP components did not match the GDP level, were omitted. Switzerland, Iceland and Norway were omitted from the donor pool sample to avoid including countries that are economically similar to those in EU. The final donor pool consists of the below 28 countries.. 政治大. Table 1: List of donor pool countries. y. ‧ 國. io. n. er. Nat. al. ‧. Korea Malaysia Morocco New Zealand North Macedonia Peru Philippines Russia Singapore Thailand Tunisia Turkey Ukraine Uruguay. 學. Albania Algeria Australia Belarus Brazil Canada Colombia Costa Rica Ecuador Hong Kong Chile Israel Japan Jordan. sit. 立. Ch. engchi. i Un. v. Source: Created by author. 18. DOI:10.6814/NCCU202000585.

(31) Data Description. Variables The choice of the outcome variables and covariates used for the SCM analysis was based on Abadie et al. (2015), complemented by several new variables to achieve the best fit of the synthetic predictors of the outcome variables in the pre-intervention period. Moreover, the same set of covariates and outcome variables was used for all four V4 countries. Outcome Variables The primary objective of this thesis is to estimate the effect of joining the EU on the GDP per capita levels of the V4 countries. Therefore, the main outcome variable is GDP per capita in Purchasing Power Parity (PPP) in constant 2011 international dollars (USD)19. The choice of the outcome variable follows the previous literature (Abadie &. 政治大. Gardeazabal, 2003; Abadie et al., 2015).. 立. Furthermore, to determine the effects on the GDP components, SCM was also. ‧ 國. 學. build for each of the GDP components. The components were calculated from the above-mentioned GDP per capita and the component share in GDP. Therefore, the. ‧. additional outcome variables are import per capita, export per capita, government spending per capita, investment per capita, private consumption per capita and lastly,. Nat. sit er. io. Treatment. y. net export per capita.. al. n. iv n C h e n gFurthermore, this year is also selected as the treatment. c h i U the V4 countries are the only in. The chosen treatment is joining the EU. The V4 countries joined the EU in 2004 and. the dataset to receive the treatment over the studied period. Due to this, countries with relatively similar economic environment, such as Romania or Croatia, had to be omitted from the donor pool as they joined the EU during the post-treatment period and thus, would provide inaccurate results. Countries that have not joined the EU, however, are strongly connected with the EU via trade agreements, were also excluded from the dataset20.. 19. In the following text, USD refers to constant 2011 international dollars in PPP. Those countries are Iceland, Norway and Switzerland. Nonetheless, they were used for the robustness check. 20. 19. DOI:10.6814/NCCU202000585.

(32) Data Description. Predictors The selection of the predictors follows the previous research papers (Abadie & Gardeazabal, 2003; Abadie et al., 2015), complemented by new variables to provide a better match of the model21. Some variables used in the above-mentioned sources were omitted as they are later used as an outcome variable 22 . Below is list of the variables used in the model. The same set of predictors was used for GDP and its components to assure the same sources of changes in the outcome. Table 2: List of predictors Age dependency ratio as % of working-age population Agriculture, value added as % of GDP. 治政大 Industry, value added as % of GDP 立 Inflation, consumer prices, annual % Human Development Index23. 學. ‧ 國. 24. Labor force participation rate, % of total population ages 15-64 Population growth, annual %. ‧. Source: Created by author. sit. y. Nat. io. al. er. Summary statistics. v. n. The tables below present summary statistics of the chosen variables for each of the. Ch. i Un. treated countries and for the donor pool. The values for inflation are described only for. engchi. years 1996-2017 as inflation in the latter estimates was averaged only for the previously mentioned years due to hyperinflation in the treated countries.. 21. Several different models with different set of predictors were tested. However, the results were not affected by those changes and therefore, were not included to the set of covariates. 22 Variables with high probability of being affected by joining the EU were also omitted. 23 Human Development Index (HDI) is an indicator composed of human development indicators, such as life expectancy, education and living standards. It ranges from 0 to 1, the higher score is assumed to the more developed countries. 24 The inflation levels are estimated only for years 1996 to 2003 due to hyperinflation that the V4 countries faced in the transition period.. 20. DOI:10.6814/NCCU202000585.

(33) Data Description. Table 3: Summary statistics of variables for Czech Republic Variable indust agri hdi pop_grow age_dep labor cpi gdpcap_2011 expcap_2011 impcap_2011 netexpcap_2011 private_conscap_2011 gov_spendcap_2011 investcap_2011. Mean 34.1513 2.6486 0.8180 0.0924 45.0728 71.6543 3.1788 24459.1500 14618.0100 14064.9600 553.0480 11933.6100 4945.7730 6988.9000. 立. Std. Dev. 0.9965 0.7712 0.0529 0.2709 3.8009 1.6221 2.9233 4839.9620 6649.8710 5833.3140 950.0359 2122.0500 891.6277 1326.9640. Min 32.6280 1.5195 0.7260 -0.3757 40.4004 69.6950 0.1187 17624.6800 6757.7100 5855.9900 -977.1596 8454.3990 3653.9730 3936.6660. 政治大 Source: Author’s calculations. Max 36.7283 4.1955 0.8880 0.8294 52.3974 76.1260 10.6984 32570.7800 25967.4100 23517.0200 2450.3910 15443.8900 6249.4780 9250.9490. ‧ 國. 學. Table 4: Summary statistics of variables for Hungary. n. Ch. engchi. y. Min 25.0363 3.0582 0.7020 -0.5164 44.8274 57.0970 -0.2276 14618.4500 3390.2760 4086.7370 -1049.2400 8150.9070 3321.7160 2418.0850. sit. Std. Dev. 0.9263 1.2086 0.0435 0.0962 1.7403 3.5831 5.8508 3934.9270 6486.3460 5738.5780 1010.5750 1820.9390 655.3797 1007.8090. er. io. al. Mean 26.2592 4.4415 0.7871 -0.2154 46.8417 62.0737 6.6294 20245.7600 13548.0200 13239.5000 308.5181 10807.4600 4388.1180 4748.5560. ‧. Nat. Variable indust agri hdi pop_grow age_dep labor cpi gdpcap_2011 expcap_2011 impcap_2011 netexpcap_2011 private_conscap_2011 gov_spendcap_2011 investcap_2011. i Un. v. Max 28.0243 7.2625 0.8410 -0.0057 51.2317 71.0820 23.4690 27031.7800 23556.3300 21586.0100 2260.7370 13387.0500 5498.0430 6176.3680. Source: Author’s calculations. 21. DOI:10.6814/NCCU202000585.

(34) Data Description. Table 5: Summary statistics of variables for Slovak Republic Variable indust agri hdi pop_grow age_dep labor cpi gdpcap_2011 expcap_2011 impcap_2011 netexpcap_2011 private_conscap_2011 gov_spendcap_2011 investcap_2011. Mean 30.1882 2.1695 0.7926 0.0966 44.0073 70.0714 4.2811 20377.1200 15204.2200 15398.0800 -193.8656 11249.7300 4080.0230 5247.8770. Std. Dev 1.8870 0.4108 0.0416 0.1327 5.2718 1.8519 3.4825 6055.3260 7775.5390 7060.3600 928.7045 3476.4810 810.9975 1332.7370. Min 27.0938 1.6188 0.7330 -0.1830 38.4572 68.2910 -0.5200 11679.6000 6219.1250 6653.5400 -1863.5470 5980.6410 3013.5520 2502.6560. 學. ‧ 國. 政治大 Source: Author’s calculations 立. Max 34.2862 3.2276 0.8540 0.3941 54.8611 76.2230 12.0358 30058.6300 28587.5300 27929.7100 1493.1700 16795.1600 5684.2290 7224.6860. Table 6: Summary statistics of variables for Poland. n. Ch. engchi. sit. y. Min 26.7242 2.2014 0.7110 -1.0443 40.2326 63.1840 -0.8741 9521.8340 2594.6150 2341.1890 -1027.8490 6690.1800 2132.9410 2223.7540. er. io. al. Std. Dev. 1.4135 0.8561 0.0475 0.2457 4.5031 2.0493 5.2137 5521.9550 3722.0570 3319.4920 597.7231 2770.2360 871.4766 981.3241. ‧. Mean 29.4760 3.1108 0.7984 -0.0132 45.6019 65.9927 4.5723 17457.9000 7246.5430 7434.1980 -187.6551 11519.2300 3426.6740 4001.3000. Nat. Variable indust agri hdi pop_grow age_dep labor cpi gdpcap_2011 expcap_2011 impcap_2011 netexpcap_2011 private_conscap_2011 gov_spendcap_2011 investcap_2011. i Un. v. Max 33.2424 5.5352 0.8680 0.3547 53.6910 69.7710 19.7950 27378.8900 14881.5500 13736.1900 1145.3610 15972.0500 4842.3850 5427.6490. Source: Author’s calculations. 22. DOI:10.6814/NCCU202000585.

(35) Data Description. Table 7: Summary statistics of variables for donor pool Variable indust agri hdi pop_grow age_dep labor cpi gdpcap_2011 expcap_2011 impcap_2011 netexpcap_2011 private_conscap_2011 gov_spendcap_2011 investcap_2011. Mean 29.1927 7.8966 0.7494 1.1456 52.6705 66.2926 7.4965 18314.2500 11445.0100 10949.3400 495.6697 10393.9700 2805.3590 4627.5990. 立. Std. Dev. 7.5660 5.3897 0.0966 1.0704 10.8953 9.4541 17.9922 14031.7300 25069.9500 22535.0400 3051.7960 6793.6360 2153.9110 3823.2990. Min 6.7172 0.0261 0.4640 -1.4745 26.9906 41.5680 -4.0094 3015.3910 241.7035 760.8411 -3619.256 2369.1550 368.0793 170.5712. 政治大 Source: Author’s calculations. Max 58.8857 36.4107 0.9370 6.0170 91.3425 81.7120 293.6788 87760.3700 157335.4000 139254.2000 22750.4300 37612.2400 9151.1030 24715.7600. ‧ 國. 學 ‧. The further explanation of the variable and the source is displayed in Appendix A: Data description. The above variables refer to: cpi = inflation, indust = industry, value added as % of GDP, agri = agriculture, value added as % of GDP, age_dep = age dependency ratio as % of working-age population, labor = labor force participation rate as % of total population ages 15-64, hdi = human development index, pop_grow = annual population growth in %, gdpcap_2011 = GDP per capita, expcap_2011 = Export per capita, impcap_2011 = Import per capita, netexpcap_2011 = Net Export per capita, private_conscap_2011 = Private Consumption per capita, gov_spendcap_2011 = Government Spending per capita, investcap_2011 = Investment per capita. GDP and GDP components are measured in constant 2011 international $, PPP.. n. er. io. sit. y. Nat. al. Ch. engchi. 23. i Un. v. DOI:10.6814/NCCU202000585.

(36) Methodology. 5 Methodology This thesis uses SCM to estimate the effects of the EU accession on the GDP per capita and the GDP components of the V4 countries. In this case, the treatment is year 2004, the EU accession date, and the treated units are the V4 countries. The main advantage of using the SCM compared to other methods lies in the ability to build a synthetic entity without receiving the treatment based on a weighted average of several different countries. The SCM method uses long-term data to estimate the development of a synthetic country that have never received the treatment. The dataset consists of a sample of J + 1 countries that are being monitored, but just one unit j = 1 in the dataset is exposed to the treatment (EU membership).The rest of the countries j = 2 to j = J + 1 are defined. 政治大 𝑡 =立 1 , … , 𝑇 periods, where 𝑡 = 1 , … , 𝑇. as the donor pool. Those are the potential candidates for creating a synthetic unit. The sample is observed over. 0. denotes the. ‧ 國. 學. pre-treatment period and 𝑡 = 𝑇0 + 1, … , 𝑇 post-treatment period. The panel needs to be balanced and must include a positive value of both, pre-intervention and. ‧. post-intervention periods.. Let 𝑌𝑖𝑡0 be the potential outcome observed for unit i at time t if the unit does not. sit. y. Nat. receive a treatment. Contrary, 𝑌𝑖𝑡1 is a potential outcome being observed for unit i at time. er. io. t if the unit receives the treatment in the period 𝑡 = 𝑇0 + 1, … , 𝑇. Assumed the treatment does not have any effect on the outcome variable in the pre-treatment period, therefore. n. al. Ch. i Un. v. 𝑌𝑖𝑡1 = 𝑌𝑖𝑡0 in the pre-treatment period. The effect of the treatment for unit i at time t is equal to αit = 𝑌𝑖𝑡1 − 𝑌𝑖𝑡0 .. engchi. Let Dit be an indicator that equals to 1 if unit i receives treatment at time t, 0 otherwise. As only the first unit is treated and only in the post-treatment period, Dit = 1 if i = 1 and t > T0, zero otherwise. The observed outcome for unit i at time t can be also expressed as 𝑌𝑖𝑡 = 𝑌𝑖𝑡0 + ∝𝑖𝑡 𝐷𝑖𝑡 . We aim to estimate (∝1𝑇0+1 , … , ∝1𝑇 ) for t > T0 as ∝1𝑡 = 𝑌𝑖𝑡1 − 𝑌𝑖𝑡0 = 𝑌𝑖𝑡 − 𝑌𝑖𝑡0 . Since 𝑌𝑖𝑡1 is observed in the dataset, to estimate ∝1𝑇 , we need to determine 𝑌𝑖𝑡0 , which represents the synthetic country that have never received the treatment.. 24. DOI:10.6814/NCCU202000585.

(37) Methodology. Suppose the potential outcome without treatment 𝑌𝑖𝑡0 is given by the below equation 𝑌𝑖𝑡0 = 𝛿𝑡 + 𝜃𝑡 𝑍𝑖 + 𝜆𝑖 µ𝑖 + 𝜀𝑖𝑡 , where 𝛿𝑡 represents the unknown common factor with constant factors loadings for the countries, 𝜃𝑡 is a (1 x r) vector of unknown parameters, 𝑍𝑖 is a (r × 1) vector of the observed covariates25 , 𝜆𝑖 is a vector (1 x F) of unobserved common factors, µ𝑖 is a (F x 1) vector of unknown factors loading, 𝜀𝑖𝑡 are unobserved transitory shocks at the unit level with zero mean for all i (Abadie et al., 2015). The SCM is described as a (J × 1) vector of weights 𝑊 = (𝑤2 … 𝑤𝑗+1 )′ where 0 ≤ 𝑤𝑗 for j = 2 ,..., J +1 and 𝑤2 + ⋯ + 𝑤𝑗+1 = 1. Each value of W symbolizes a possible synthetic control26. Abadie et al. (2015) show the formula for the estimate of the effect is expressed as the difference between the obtained outcome for the treated. 政治大. unit and the estimated outcome for the synthetic control variable at that period as in the equation below:. 立. 𝐽+1. ‧ 國. 學. ̂ 1𝑡 = 𝑌1𝑡 − ∑ 𝑤𝑗∗ 𝑌𝑗𝑡 ∝ 𝑗=2. ‧. 𝐽+1 ∗ 0 We want to select 𝑊 ∗ such that ∑𝐽+1 𝑗=2 𝑤𝑗 ∗ 𝑍𝑗 = 𝑍1 and that ∑𝑗=2 𝑤𝑗 µ𝑗 = µ1 so that 𝑌𝑖𝑡. is an unbiased estimator. Let X1 be a vector (k × 1) that contains the pre-treatment. sit. y. Nat. characteristics of the treated unit and let X0 be the (k × J) matrix that collects values. al. er. io. from the donor pool of the same variable. Therefore, the vector 𝑋1 − 𝑋0 𝑊27. n. characterizes the difference between the synthetic and real values in the pre-treatment. Ch. i Un. v. period28. We aim to select W* that minimize the difference between the two variables.. engchi. We want to select the weights that minimize the root mean square prediction error (RMSPE) 29 . The countries disponing with most similar characteristics to the pre-treatment treated unit will be chosen and given values depending on their similarities. In total, those values sum to 1. The choice of V influences the means square of the estimator. 25. Not affected by the intervention. Possible weighted average of the control units in the donor pool. 27 Subject to 0 ≤ 𝑤2 ,...,0 ≤ 𝑤𝑗+1 , 𝑤2 + ⋯ + 𝑤𝑗+1 = 1 28 The difference between X and X W is measured as ‖𝑋 − 𝑋 𝑊‖𝑣 = √(𝑋 − 𝑋 𝑊)′𝑉(𝑋 − 𝑋 𝑊), 1 0 1 0 1 0 1 0 where V is (k x k) symmetric and positive semidefinite matrix (Abadie et al., 2015). V is chosen to minimize the mean squared prediction error (MSPE) of the outcome variable for the treated country in the pre-intervention period, meaning the mean of squared deviation between the real outcome variable for the treated country and the synthetic values. 26. 29. 1. 𝑇0 𝑗+1 𝑅𝑀𝑆𝑃𝐸 = √ ∑𝑡=1 (𝑌1𝑡 − ∑𝑗=2 𝑤𝑗∗ 𝑌𝑗𝑡 ). 2. 𝑇0. 25. DOI:10.6814/NCCU202000585.

(38) Methodology. The statistical inference of comparative studies in general is challenging due to small sample of data, lack of randomization and inability to use traditional statistical inference (Abadie et al., 2015). The main tools to test robustness of the results are either in-time30 or in-space placebo effects31. P-values can be estimated to show the probability of obtaining larger or the same results for other units in the donor pool. By using the pre-treatment RMSPE, the quality of the donor pool match can be tested by using standard two-sided p-values constructed by estimating in-space placebo effect of each donor pool unit32. The effects greater or equal to the estimated effect for the treated unit is calculated as 𝑝 − 𝑣𝑎𝑙𝑢𝑒𝑡𝑠𝑡𝑑 = ̂ |𝛼. 𝑡,𝑝𝑜𝑠𝑡. |. 𝑡,𝑝𝑜𝑠𝑡. ̂ 1𝑡 |𝛼. |. 𝑗𝑡 33 Pr (𝑅𝑀𝑆𝑃𝐸 (Kumar & Liang, 2018). Following Abadie et al., (2015), 𝑝𝑟𝑒 ≥ 𝑝𝑟𝑒 ) 𝑅𝑀𝑆𝑃𝐸 1𝑡. 𝑗𝑡. countries with five times higher pre-treatment RMSPE were omitted from the donor. 政治大. pool.. 立. ‧. ‧ 國. 學. n. er. io. sit. y. Nat. al. Ch. engchi. i Un. v. 30. In-time placebo effects are feasible for estimates with long pre-treatment period. The treatment is applied to different date and if the retrieved results produce large effects, the credibility of the original estimates is weakened. 31 This method applies the treatment to the donor pool countries and if the retrieved effect is higher for any other unit in the donor pool, the credibility is weakened. 32. 1. ̂ 𝑡,𝑝𝑟𝑒 2 ) and the post-treatment The pre-treatment RMSPE is constructed as 𝑅𝑀𝑆𝑃𝐸 𝑝𝑟𝑒 = √ ∑𝑡≤𝑇0(∝ 𝑇 0. RMSPE is constructed as 𝑅𝑀𝑆𝑃𝐸 33. 𝑡,𝑝𝑜𝑠𝑡. 𝛼̂𝑗𝑡. 𝑝𝑟𝑒. and 𝑅𝑀𝑆𝑃𝐸𝑗𝑡. 𝑝𝑜𝑠𝑡. 1. ̂ 𝑡,𝑝𝑜𝑠𝑡 2 ) ∑𝑇0+1≤𝑡≤𝑇(∝ = √ 𝑇−𝑇 0. characterize the donor pool countries estimates.. 26. DOI:10.6814/NCCU202000585.

(39) Preliminary results. 6. Preliminary results. The below preliminary results will help to construct assumptions about the expected development of the outcome variables. Overall, in terms of GDP per capita, the Czech Republic is the richest country in the sample, while Poland’s values are the lowest. After joining the EU, Slovakia reached the highest level of economic growth. Hungary seems to have achieved much less from the EU membership than the other V4 countries.. GDP per capita The graph below presents the values of the GDP per capita for the V4 countries over the studied period, where the Czech Republic reaches the highest level of GDP per capita, while Poland and Hungary reached the lowest values in the last years.. 政治大 overgrew the Hungarian立 values of GDP per capita. Moreover, the gap between the. Furthermore, the Slovak GDP per capita grew much faster after year 2004 and its values. ‧ 國. 學. Slovak and Czech GDP per capita seems to close over the last years. Furthermore, Czech Republic, Slovakia and Poland show strong growth after joining the EU, while the line for Hungary show much less strong positive trend.. ‧ y. Nat. n. al. er. io. sit. Figure 1: GDP per capita in the treated countries and EU15. Ch. engchi. i Un. v. Source: World Bank, author’s calculations. 27. DOI:10.6814/NCCU202000585.

(40) Preliminary results. The above graph also shows that even though the Polish gross GDP value is a lot higher compared to the other countries in the V4 group, the Polish GDP per capita was the lowest over the studied period. During the studied years, Poland reached the lowest value of GDP per capita in 1991 with GDP per capita reaching only 9 522 USD. Meanwhile, the highest GDP per capita of 32 571 USD was reached in 2017 by the Czech Republic. The growth of the GDP per capita largely decreased due to the economic crisis in all V4 countries, except for Poland. The above graph also shows the average GDP per capita of the old EU states34. The average GDP per capita for the EU15 is well above the values for the V4 countries and the graph shows some, but low, convergence level after V4 joined the EU. The V4 countries (except of Hungary) experienced constant growth of GDP per capita and therefore, the results of the SCM for GDP per capita are expected to be positive35.. 政治大. Overall, the mean GDP per capita of the studies countries is 20 637 USD with. 立. standard deviation of 5 659 USD. From the V4 countries, Slovak Republic reached the. ‧ 國. 學. highest GDP growth and in 2017, the GDP growth reached almost 11%. Nonetheless, the positive effect was offset in 2009 by the global economic crisis when Slovakia’s. ‧. GDP per capita growth fell nearly to -6%.. Nat. sit. y. Trade balance per capita. io. er. When looking at import and export per capita over the studied period, Poland is much less economically open than the other V4 countries, which is caused by the greater. n. al. Ch. i Un. v. internal market compared to the others. The studied countries experienced a slump due. engchi. to the global economic crisis in both, exports and imports. Both variables follow the same pattern. The graph shows that Slovakia reached in the last years the highest export per capita, which can be associated with the EU accession and also the latter eurozone accession. Based on the below graphs, we can assume that the EU participation caused a large increase of the trade balance as the new EU countries joined the single market and benefited from trade with the EU countries.. 34 35. The GDP per capita is averaged for the old EU15. The results for Hungary are expected to be negative.. 28. DOI:10.6814/NCCU202000585.

(41) Preliminary results. Figure 2: Trade balances for the V4 countries. 政治大. 立. ‧. ‧ 國. 學. Nat. n. al. er. io. sit. y. Source: World Bank, author’s calculations. Government spending per capita. Ch. engchi. i Un. v. Overall, government spending per capita experienced an increasing trend in all the studied countries with a slump in 2009. Just as for the previous GDP components, Poland shows the most stable growth of government spending. The graph also shows a large drop in government spending per capita in Hungary few years after joining the EU, which is however due to a change in the strategy of the government to decrease the budget deficit of Hungary by cutting on government investment. After joining the EU, government spending increased in each of the V4 countries. Arpaia and Turrini (2008) found proportional relation between the growth of government spending per capita and output level. Therefore, we expect the government spending per capita to follow the same trend as for GDP per capita, however, the positive effect is expected to be smaller. The positive effect probably will not be experienced by Hungary as due to the large slump after the economic crisis.. 29. DOI:10.6814/NCCU202000585.